Cursor rule

1---
2description: "Create new bigpowers skills with proper structure, progressive disclosure, and bundled resources. Use when user wants to create, write, or build a new skill for the bigpowers lifecycle."
3alwaysApply: false
4---
5 
6# Craft Skill
7 
8> **HARD GATE** — Do NOT name a skill without a two-word verb-noun pair. Do NOT merge a new skill without running `sync-skills.sh`. Generated `.cursor/rules/` and `.gemini/` artifacts MUST match the source SKILL.md.
9 
10## CSO Description Discipline (e45s02)
11 
12The YAML `description` is the **Catalog Selection Object** — the only field agents see when picking a skill.
13 
14| Rule | Limit |
15|------|-------|
16| Max length | 1024 characters |
17| Voice | Third person |
18| Content | Capability + `Use when …` triggers only |
19| Forbidden | Workflow steps, phase chains, numbered lists, `→ verify:`, HARD GATE prose |
20 
21Move process detail into the SKILL.md body or REFERENCE.md — never into `description`.
22 
23## Agentic STE body discipline (e79s02)
24 
25Skill-body instructional prose MUST follow [AGENTIC-STE.md](../../docs/AGENTIC-STE.md).
26 
27| Rule | Limit |
28|------|-------|
29| Sentence length | ≤20 words per instruction sentence |
30| Voice | Imperative, active |
31| Directive terms | MUST, MUST NOT, NEVER, ALWAYS, DO, DO NOT |
32| Banned modals | should, might, could, may, consider, try, generally, typically |
33| Scope | SKILL.md body only — not YAML `description`, not `terse-mode` output |
34 
35> **HARD GATE** — Do NOT merge a new or edited skill until `bash scripts/validate-agentic-ste.sh --strict skills/<name>/SKILL.md` exits 0. Fix violations before `sync-skills.sh`.
36 
37## Process
38 
391. **Gather requirements** — ask user about:
40   - What task/domain does the skill cover?
41   - Which use cases must the skill handle?
42   - Does it need executable scripts or just instructions?
43   - Any reference materials to include?
44   - What specs/ output does it produce (if any)?
45 
462. **Verify Principles** — Ensure the skill aligns with [PRINCIPLES.md](../../docs/PRINCIPLES.md):
47   - Is it atomic (verb-noun)?
48   - Is it "deep" (simple interface, complex internal logic)?
49   - Does it include Hard Gates?
50   - Is it verifiable with a `.feature` file?
51 
523. **Draft the skill** — create:
53   - SKILL.md with concise instructions (see [REFERENCE.md](REFERENCE.md) for template)
54   - Additional reference files if content exceeds 100 lines
55   - Utility scripts if deterministic operations needed
56 
57   **Auto-skill from library README:** When user provides a library README or API docs URL, extract triggers and HARD GATEs.
58   Draft verify commands and specs/ output into SKILL.md. Do NOT invent APIs not in the source.
59 
604. Add `model:` frontmatter (`haiku` | `sonnet` | `opus`) per [model-profiles.md](../../docs/references/model-profiles.md).
61 
62> **STREAM CONTINUITY** — When writing file content, output in continuous chunks of ~200 lines. Do not pause. Continue immediately until complete. If you need time, emit a placeholder comment rather than going silent.
63 
645. **Review with user** — present draft and ask:
65   - Does this cover your use cases?
66   - Anything missing or unclear?
67   - Does any section need more or less detail?
68 
696. **Completion-honesty gate (HARD GATE — e45s02)** — Before declaring done:
70   - Run `bash scripts/validate-skill-description.sh skills/<name>/SKILL.md` — must exit 0
71   - Run `bash scripts/validate-agentic-ste.sh --strict skills/<name>/SKILL.md` — must exit 0 (e79s02)
72   - Run `bash scripts/sync-skills.sh` — must complete without error
73   - Run `bash scripts/run-skill-verify.sh <name>` if the skill defines a verify command
74   - Show terminal output for each — narration without evidence is rejected
75 
76## Naming Rules
77 
78Every skill name must be a **two-word verb-noun pair**. See [REFERENCE.md](REFERENCE.md) for full rules, examples, and documented exceptions.
79 
80## specs/ Output
81 
82If the skill produces written output, it goes in `specs/` at the project root. Document the output file path in the skill body and in CONVENTIONS.md's output files table.
83 
84## Review Checklist
85 
86After drafting, verify:
87 
88- [ ] Name is a two-word verb-noun pair (or follows grill-me exception)
89- [ ] Description < 1024 chars, triggers only, no workflow-summary leakage
90- [ ] Description includes triggers ("Use when...")
91- [ ] SKILL.md under 100 lines
92- [ ] No time-sensitive info
93- [ ] Consistent terminology with CONVENTIONS.md
94- [ ] specs/ output documented if applicable
95- [ ] `validate-skill-description.sh` exits 0
96- [ ] `validate-agentic-ste.sh --strict` exits 0 (e79s02)
97- [ ] `sync-skills.sh` run to propagate to Cursor/Gemini
98- [ ] `bash scripts/validate-skill-catalog.sh` passes for the new skill (HARD GATE — completion honesty)
99 
100> **HARD GATE** — Do NOT declare the skill done until `bash scripts/validate-skill-catalog.sh --strict --skill <name>` exits 0. Validator enforces verb-noun name, HARD GATE block, description ≤1024 chars, and `→ verify:` command.
101 
102## Verify
103 
104→ verify: `bash scripts/validate-skill-catalog.sh --strict --skill craft-skill && bash scripts/validate-skill-description.sh skills/craft-skill/SKILL.md`
105 
106---
107 
108# Craft Skill — Reference
109 
110## Naming Rules (full)
111 
112Every skill name must be a **two-word verb-noun pair**:
113- First word: a verb (survey, model, define, develop, audit…)
114- Second word: a noun from PMBOK 6 / Agile vocabulary (context, domain, language, tdd, code…)
115- Pronounceable in any language, searchable, no noise words, no encodings
116- Exception precedent: `grill-me` — kept for recognizability
117 
118Good: `survey-context`, `audit-code`, `validate-fix`
119Bad: `context-surveyor`, `code-auditing-skill`, `fix-validator`
120 
121Any new naming exception requires an entry in CONVENTIONS.md before the skill is published.
122 
123## Skill Structure
124 
125```
126skill-name/
127├── SKILL.md           # Main instructions (required)
128├── REFERENCE.md       # Detailed docs (if needed)
129├── EXAMPLES.md        # Usage examples (if needed)
130└── scripts/           # Utility scripts (if needed)
131    └── helper.sh
132```
133 
134## SKILL.md Template
135 
136```md
137---
138name: skill-name
139description: Brief description of capability. Use when [specific triggers].
140---
141 
142# Skill Name
143 
144## Quick start
145 
146[Minimal working example]
147 
148## Workflows
149 
150[Step-by-step processes with checklists for complex tasks]
151 
152## Advanced features
153 
154[Link to separate files: See [REFERENCE.md](REFERENCE.md)]
155```
156 
157## Description Requirements
158 
159The description is **the only thing your agent sees** when deciding which skill to load.
160 
161**Format**:
162- Max 1024 chars
163- Write in third person
164- First sentence: what it does
165- Second sentence: "Use when [specific triggers]"
166 
167**Good example**:
168```
169Investigate a bug by exploring the codebase to find root cause, then write a TDD-based fix plan to specs/bugs/BUG-*.md. Use when user reports a bug, wants to investigate a problem, or mentions "triage".
170```
171 
172## When to Add Scripts
173 
174Add utility scripts when:
175- Operation is deterministic (validation, formatting)
176- Same code would be generated repeatedly
177- Errors need explicit handling
178 
179## When to Split Files
180 
181Split into separate files when:
182- SKILL.md exceeds 100 lines
183- Content has distinct domains
184- Advanced features are rarely needed
185 
186## sync-skills.sh Propagation
187 
188After adding a new skill directory with SKILL.md, run `scripts/sync-skills.sh` from the bigpowers repo root. This automatically generates:
189- `.cursor/rules/<name>.mdc` — for Cursor
190- `.gemini/extensions/bigpowers/skills/<name>/SKILL.md` — Agent Skill
191- `.gemini/extensions/bigpowers/commands/<name>.toml` — Slash Command
192- `.gemini/extensions/bigpowers/commands/prompts/<name>.md` — Command Prompt
193- Updated `gemini-extension.json`
194 
195verify: `bash scripts/sync-skills.sh 2>&1 | grep "skills synced"`
196